The Customer Confidential Podcast

The Human Side of Uber’s Digital Customer Experience

Uber’s Troy Stevenson talks about the art and science of delivering great experiences in a hypergrowth environment.


The Human Side of Uber’s Digital Customer Experience

Most of us think about Uber as a true digital disrupter. Almost everything you need to do can be achieved directly through Uber’s mobile app. In fact, Uber often serves as one of the best illustrations of what a “mobile-first” business model looks like.

Yet, there is a very large human side to Uber. When a driver, a rider or what Uber refers to as “an eater” (someone ordering food through Uber Eats) needs help resolving something that can’t be addressed purely in-app, they contact Community Operations. That’s an organization of many thousands of people working in local markets around the world to serve Uber’s customer base via in-app chat, on the phone or even in person.

Troy Stevenson, global head of Community Operations, has the big job of ensuring that all these customer interactions are completed both with high quality and in a cost-effective way. Uber supports millions of customers every day. Community Operations handles everything from the mundane challenges of reuniting riders with belongings they left in a car to resolving more complex issues, like disputes between riders and drivers. But Troy’s organization goes far beyond just reacting to or resolving customer service problems. It also supports getting new drivers set up to drive for Uber in the first place, and helps ensure they retain them. Troy’s operation includes hundreds of locations around the world, called Greenlight Hubs, that can set drivers up to drive for Uber or provide ongoing help if they need it. Community Operations also supports Uber’s other ventures, from Uber Eats to bikes and scooters. And Troy’s organization is responsible for everything related to safety, so it investigates incidents or accidents and works to make sure that Uber is as safe as possible.

It’s also a relatively new organization. Less than a decade ago, almost everyone in the company—all the way to the CEO–spent time handling customer issues. That model certainly kept the voice of the customer close to every employee, but Uber’s rapid growth made that model untenable, and it wasn’t a recipe for consistently delivering a positive experience.

Formed about four years ago, Community Operations is still adapting and growing. In our conversation, Troy describes how he helped Uber navigate the growing pains that resulted when the company expanded far faster than anyone had anticipated. And he also has something many customer-focused executives would envy: vast amounts of real-time customer data, most of it in modern, accessible databases. In this podcast, he shares some of the ways that he makes sure that data gets turned into action, and talks about why that’s both a science and an art.

You can listen to my conversation with Troy on iTunesStitcher or your podcast provider of choice, or through the audio player below.

Content requires cookies

This content requires non-essential cookies for the functionality to work. 
To view this content, please consent to the use of cookies. You can learn more by reading our Cookie Policy.


In the following excerpt, Troy, a veteran of customer-focused roles at both Charles Schwab and eBay, talks about how the vast flows of customer feedback data make his job different than it was at previous companies, even though the end goal is exactly the same.

Troy Stevenson: One big thing that we’ve done is reaching out to customers that have told us they had a bad experience to try to really understand what happened and how can we do better.

So it’s a little bit different than the pure data-oriented look at it feedback logs.

Rob Markey: It’s a more qualitative understanding.

Troy Stevenson: So we tied it to low the survey scores. So we’ll have something where we continually get, on the various post-interaction surveys, either low scores with the agent or low scores with Uber. We’ve got a team that will rotate from topic to topic. They’ll pull a robust-enough sample that we feel like we’ll get, you know, a meaningful amount of data, and they’ll really try to dig into what happened, why did it happen, and then step back and try to codify it, write up a report, and then that’s something we can syndicate across the team. And that’s been a big help at getting our customer satisfaction scores up.

We were talking before about just eliminating root cause defects. This effort is more about, for those defects that still remain, how can we make sure that we’re mitigating them and managing them in a way that results in a good customer experience.

Rob Markey: This is a great example of the spirit of follow-up calls in the Net Promoter System®, because you’re doing it in order to learn more.

Troy Stevenson: Yeah, it’s more than just the spirit. It is what we’re doing!


Rob Markey: OK, OK [laughter]. And so the express intent is learning enough to do something about it. And I assume that in the process of doing that, the folks who are doing these calls also have developed skills for making the customer really feel heard and feel like they’re making a difference by spending the time with them.

Troy Stevenson: We did a similar thing at Schwab where we called detractors back.

Rob Markey: Yes, we made a video about you guys because you were one of the early [Net Promoter® practitioners].

Troy Stevenson: It was just something we felt so passionately about. The customer’s not happy, we’re not going to leave it at that. We’re going to fix it.

Now, you couldn’t do that at Uber, because you would need a team of thousands making those calls just given the size. And so we’ve kind of pivoted on that strategy—it’s the same philosophy with a different approach.

Rob Markey: I like this, because when you’re the scale that you guys are at, if you can’t call every customer whose relationship seems like it might be at risk, then focus on something, some issue, that puts a lot of relationships at risk.

Troy Stevenson: It was a scalability thing. So our mindset at Schwab was we’re going to call them all back because those calls are going to have an impact on a big enough subset of the customers to actually impact the business. We could never do that at Uber. The point of the calls at Uber is—although we’re only going to talk to a tiny subset of the people that had that issue—we’re going to be able to expand the insights we get from those calls to everybody else. And so then we will rotate topic by topic. And so in a business where, at our scale, you couldn’t hire enough people to call everybody that has an issue, you can still have an impact on all those people by talking to enough of them where you learn the root causes, learn the fixes and then systematically apply them.

Net Promoter®, Net Promoter System®, Net Promoter Score® and NPS® are registered trademarks of Bain & Company, Inc., Fred Reichheld and Satmetrix Systems, Inc.


Want to hear more from today's loyalty leaders?

Explore more episodes of The Customer Confidential Podcast.


Want to continue the conversation?

We offer unparalleled analytic and organizational tools for the Net Promoter System. Together, we can create an enduring customer-centric culture.